T lymphocytes are formed in the bone marrow, migrate to and mature in the thymus and then enter the peripheral blood and lymphatic circulation. T lymphocytes can be phenotypically subdivided into several distinct types of cells including: helper T cells, suppressor T cells, and cytotoxic T cells. T lymphocytes, unlike B lymphocytes, do not produce antibody molecules, but express a heterodimeric cell surface receptor that can recognize peptide fragments of antigenic proteins that are attached to proteins of the major histocompatibility complex (MHC) expressed on the surfaces of target cells; see, e.g., Abbas, A. K., Lichtman, A. H., and Pober, J. S., Cellular and Molecular Immunology, 1991, esp. pages 15-16.
T lymphocytes that can be expanded according to the present invention are of particular interest in the context of cellular "immunotherapy". As used herein, cellular immunotherapy refers to any of a variety of techniques involving the introduction of cells of the immune system, especially T lymphocytes, into a patient to achieve a therapeutic benefit. Such techniques can include, by way of illustration, "immuno-restorative" techniques (involving, e.g., the administration of T cells to a patient having a compromised immune system); "immuno-enhancing" techniques (involving, e.g., the administration of T cells to a patient in order to enhance the ability of that patient's immune system to avoid or combat a cancer or a pathogen such as a virus or bacterial pathogen); and "immuno-modulating" techniques (involving, e.g., the administration of T cells to a patient in order to modulate the activity of other cells of the patient's immune system, such as in a patient affected by an autoimmune condition).
Cytotoxic T lymphocytes (CTLs) are typically of the CD3+, CD4-, CD8+ phenotype and lyse cells that display fragments of foreign antigens associated with class I MHC molecules on their cell surfaces. CTLs that are CD3+, CD4+, CD8- have also been identified. Target cells for CTL recognition include normal cells expressing antigens after infection by viruses or other pathogens; and tumor cells that have undergone transformation and are expressing mutated proteins or are over-expressing normal proteins.
Most "helper" T cells are CD3+, CD4+, CD8-. Helper T cells recognize fragments of antigens presented in association with class II MHC molecules, and primarily function to produce cytokines that amplify antigen-specific T and B cell responses and activate accessory immune cells such as monocytes or macrophages. See, e.g., Abbas, A. K., et al., supra. Helper T cells can also participate in and/or augment cytolytic activites.
In addition to conventional helper T cells and cytolytic or "killer" T cells, it will also be useful to be able to rapidly expand other T cell populations. For example, T cells expressing the gamma/delta T cell receptor represent a relatively small portion of the human T cell population, but are suspected to play a role in reactivity to viral and bacterial pathogens as well as to tumor cells (see, e.g., W. Haas et al. 1993. Annu. Rev. Immunol. 11:637). Another T cell population of potential clinical importance is the population of CD1-restricted T cells. CD1 is an MHC-like molecule that shows limited polymorphism and, unlike classical MHC molecules which "present" antigenic peptides, CD molecules bind lipoglycans and appear to be important in the recognition of microbial antigens (see, e.g., P. A. Sieling et al. 1995. Science 269:227; and E. M. Beckman et al. 1994. Nature 372:691).
T lymphocytes are thus key components of the host immune response to viruses, bacterial pathogens and to tumors. The significance of properly functioning T cells is made quite clear by individuals with congenital, acquired or iatrogenic T cell immunodeficiency conditions (e.g., SCID, BMT, AIDS, etc.) which can result in the development of a wide variety of life-threatening infections or malignancies. Persons with diseases that are related to a deficiency of immunologically-competent T lymphocytes, or persons with conditions that can be improved by administering additional T lymphocytes, can thus be benefited by cellular immunotherapies, as referred to above. T cells for use in such therapies can be derived from the immunodeficient host, or from another source (preferably a compatible donor). The latter source is of course especially important in situations in which an immunodeficient host has an insufficient number of T cells, or has T cells that are insufficiently effective. In either case, it is difficult to obtain sufficient numbers of T cells for effective administration; and thus target T cells must first be grown to large numbers in vitro before administration to a host.
After undergoing such cellular immunotherapy, hosts that previously exhibited, e.g., inadequate or absent responses to antigens expressed by pathogens or tumors, can express sufficient immune responses to become resistant or immune to the pathogen or tumor.
Adoptive transfer of antigen-specific T cells to establish immunity has been demonstrated to be an effective therapy for viral infections and tumors in animal models (reviewed in Greenberg, P. D., Advances in Immunology (1992)). For adoptive immunotherapy to be effective, antigen-specific T cells usually need to be isolated and expanded in numbers by in vitro culture, and following adoptive transfer such cultured T cells must persist and function in vivo. For treatment of some human diseases, the use in immunotherapy of cloned antigen-specific T cells which represent the progeny of single cells, offers significant advantages because the specificity and function of these cells can be rigorously defined and precise dose:response effects readily evaluated. Riddell et al. were the first to adoptively transfer human antigen-specific T cell clones to restore deficient immunity in humans. Riddell, S. R. et al., "Restoration of Viral Immunity in Immunodeficient Humans by the Adoptive Transfer of T Cell Clones", Science 257:238-240 (1992). In that study, Riddell et al. used adoptive immunotherapy to restore deficient immunity to cytomegalovirus in allogeneic bone marrow transplant recipients. Cytomegalovirus-specific CD8+ cytotoxic T cell clones were isolated from three CMV seropositive bone marrow donors, propagated in vitro for 5 to 12 weeks to achieve numerical expansion of effector T cells, and then administered intravenously to the respective bone marrow transplant (BMT) recipients. The BMT recipients were deficient in CMV-specific immunity due to ablation of host T cell responses by the pre-transplant chemoradiotherapy and the delay in recovery of donor immunity commonly observed after allogeneic bone marrow transplant (Reusser et al. Blood, 78:1373-1380, 1991). Riddell et al. found that no toxicity was encountered and that the transferred T cell clones provided these immunodeficient hosts with rapid and persistent reconstitution of CD8+ cytomegalovirus-specific CTL responses.
Riddell et al. (J. Immunology, 146:2795-2804, 1991) used the following procedure for isolating and culturing the CD8+ CMV-specific T cell clones: peripheral blood mononuclear cells (PBMCs) derived from the bone marrow donor were first cultured with autologous cytomegalovirus-infected fibroblasts to activate CMV-specific CTL precursors. Cultured T cells were then restimulated with CMV-infected fibroblasts and the cultures supplemented with .gamma.-irradiated PBMCs. 2-5 U/ml of interleukin-2 (IL-2) in suitable culture media was added on days 2 and 4 after restimulation to promote expansion of CD8+ CTL (Riddell et al., J. Immunol., 146:2795-2804, 1991). To isolate T cell clones, the polyclonal CD8+ CMV-specific T cells were plated at limiting dilution (0.3-0.6 cells/well) in 96-well round bottom wells with either CMV-infected fibroblasts as antigen-presenting cells (Riddell, J. Immunol., 146:2795-2804, 1991); or anti-CD3 monoclonal antibody to mimic the stimulus provided by antigen-presenting cells. (Riddell, J. Imm. Methods, 128:189-201, 1990). Then, .gamma.-irradiated peripheral blood mononuclear cells (PBMC) and EBV-transformed lymphoblastoid cells (LCL) were added to the microwells as feeder cells. Wells positive for clonal T cell growth were evident in 10-14 days. The clonally derived cells were then propagated to large numbers initially in 48- or 24-well plates and subsequently in 12-well plates or 75-cm.sup.2 tissue culture flasks. T cell growth was promoted by restimulation every 7-10 days with autologous CMV-infected fibroblasts and .gamma.-irradiated feeder cells consisting of PBMC and LCL, and the addition of 25-50 U/ml of IL-2 at 2 and 4 days after restimulation.
A major problem that exists in the studies described above, and in general in the prior art of culturing T cells, is the inability to grow large quantities of human antigen-specific T cell clones in a timely fashion. It is not known if the slow growth of T cells in culture represents an inherent property of the cell cycle time for human lymphocytes or the culture conditions used. For example, with the culture method used in the CMV adoptive immunotherapy study described above, three months were required to grow T cells to achieve the highest cell dose under study which was 1.times.10.sup.9 T cells/cm.sup.2. This greatly limits the application of adoptive immunotherapy for human viral diseases and cancer since the disease process may progress during the long interval required to isolate and grow the specific T cells to be used in therapy. Based on extrapolation from animal model studies (reviewed in Greenberg, P. D., Advances in Immunology, 1992), it is predicted that in humans doses of antigen-specific T cells in the range of 10.sup.9 -10.sup.10 cells may be required to augment immune responses for therapeutic benefit.
However, rapidly expanding antigen-specific human T cells in culture to achieve such high cell numbers has proven to be a significant obstacle. Thus, with the exception of the study by Riddell et al., supra, (in which several months were taken to grow a sufficient number of cells) studies of adoptive immunotherapy using antigen-specific T cell clones have not been performed. The problem of producing large numbers of cells for adoptive immunotherapy was identified in U.S. Pat. No. 5,057,423. In this patent, a method for isolating pure large granular lymphocytes and a method for the expansion and conversion of these large granular lymphocytes into lymphokine activated killer (LAK) cells is described. The methods are described as providing high levels of expansion, i.e. up to 100-fold in 3-4 days of culture. Although LAK cells will lyse some types of tumor cells, they do not share with MHC-restricted T cells the properties of recognizing defined antigens and they do not provide immunologic memory. Moreover, the methods used to expand LAK cells, which predominantly rely on high concentrations of IL-2 do not efficiently expand antigen-specific human T cells (Riddell et al., unpublished); and those methods can render T cells subject to programmed cell death (i.e. apoptosis) upon withdrawal of IL-2 or subsequent stimulation via the T cell receptor (see the discussion of the papers by Lenardo et al, and Boehme et al., infra). Earlier methods that relied on the use of lectins, such as concanavalin A or phytohemagglutinin (see, e.g., Van de Griend et al., Transplantation 38: 401-406 (1984), and Van de Griend et al., J. Immunol. Methods 66: 285-298 (1984)), are even less satisfactory because the use of such non-specific stimultory lectins tends to induce a number of phenotypic changes in the stimulated cells that make them quite different from T cells stimulated via the CD3 receptor.
The inability to culture antigen-specific T cell clones to large numbers has in part been responsible for limiting adoptive immunotherapy studies for human diseases such as cancer (Rosenberg, New Engl. J. Med., 316:1310-1321, 1986; Rosenberg, New Engl. J. Med., 319:1676-1680, 1988) and HIV infection (Ho M. et al., Blood 81:2093-2101, 1993) to the evaluation of activated polyclonal lymphocyte populations with poorly defined antigen specificities. In such studies, polyclonal populations of lymphocytes are either isolated from the blood or the tumor filtrate and cultured in high concentrations of the T cell growth factor IL-2. In general, these cells have exhibited little if any MHC-restricted specificity for the pathogen or tumor and in the minority of patients that have experienced therapeutic benefit, it has been difficult to discern the effector mechanism involved. Typically, adoptive immunotherapy studies with non-specific effector lymphocytes have administered approximately 2.times.10.sup.10 to 2.times.10.sup.11 cells to the patient. (See, e.g., U.S. Pat. No. 5,057,423, at column 1, lines 40-43).
The development of efficient cell culture methods to rapidly grow T lymphocytes will be useful in both diagnostic and therapeutic applications. In diagnostic applications, the ability to rapidly expand T cells from a patient can be used, for example, to quickly generate sufficient numbers of cells for use in tests to monitor the specificity, activity, or other attributes of a patient's T lymphocytes. Moreover, the capability of rapidly achieving cell doses of 10.sup.9 -10.sup.10 cells will greatly facilitate the applicability of specific adoptive immunotherapy for the treatment of human diseases.
There are several established methods already described for culturing cells for possible therapeutic use including methods to isolate and expand T cell clones. Typical cell culture methods for anchorage-dependent cells, (i.e., those cells that require attachment to a substrate for cell proliferation) are limited by the amount of surface area available in culture vessels used (i.e., multi-well plates, petri dishes, and culture flasks). For anchorage-dependent cells, the only way to increase the number of cells grown is generally to use larger vessels with increased surface area and/or use more vessels. However, hematopoietic cells such as T lymphocytes are anchorage-independent. They can survive and proliferate in response to the appropriate growth factors in a suspension culture without attachment to a substrate. Even with the ability to grow antigen-specific lymphocytes in a suspension culture, the methods reported to date have not consistently produced rapid numerical expansion of T cell clones. For example, in a study of T cells conducted by Gillis and Watson, it was found that T cells cultured at low densities, i.e., 5.times.10.sup.3 to 1.times.10.sup.4 cell/ml in over a seven day period and eventually reached a saturation density of 3-5.times.10.sup.5 cells/ml. Gillis, S. and Watson, J. "Interleukin-2 Dependent Culture of Cytolytic T Cell Lines", Immunological Rev., 54:81-109 (1981). Furthermore, Gillis and Watson also found that once cells reached this saturation concentration, the cells would invariably die. Gillis et al., id.
Another study reported three different methods for establishing murine T lymphocytes in long-term culture. Paul et al., reported that the method most widely used is to grow T lymphocytes from immunized donors for several weeks or more in the presence of antigen and antigen-presenting cells (APCs) to provide the requisite T cell receptor signal and co-stimulatory signals, and with the addition of exogenous growth factors before attempting to clone them, Paul, W. E., et al., "Long-term growth and cloning of non-transformed lymphocytes", Nature, 294:697-699, (1981). T cells specific for protein antigens are then cloned by limiting dilution with antigen and irradiated spleen cells as a source of APCs. A second method involved growing T cells as colonies in soft agar as soon as possible after taking the cells from an immunized donor. The T cells were stimulated in an initial suspension culture with antigen and a source of APCs, usually irradiated spleen cells. In this second approach, it was found that, after 3 days, the cells were distributed in the upper layer of a two-layer soft agar culture system. The colonies were picked from day 4 to 8 and then expanded in long-term cultures. The third approach involved selecting cells for their functional properties rather than their antigenic specificity and then growing them with a series of different irradiated feeder cells and growth factor containing supernatants. Paul, W. E. et al., "Long-term growth and cloning of non-transformed lymphocytes", Nature, 294:697-699, (1981). It is apparent that with each of these methods, it is not possible to expand individual T cell clones from a single cell to 10.sup.9 -10.sup.10 cells in a timely manner. Thus, despite the ability to clone antigen-specific T cells, and convincing evidence of the therapeutic efficacy of T cell clones in accepted animal models, the technical difficulty in culturing human T cells to large numbers has impeded the clinical evaluation and application of cellular immunotherapeutic procedures.
Yet another concern with cultured T cells is that they must remain capable of functioning in vivo in order to be useful in immunotherapeutic procedures. In particular, it has been observed that antigen-specific T cells which were grown long term in culture in high concentrations of IL-2 may develop cell cycle abnormalities and lose the ability to return to a quiescent phase when IL-2 is withdrawn. In contrast, the normal cell cycle consists of four successive phases: mitosis (or "M" phase) and three phases which make up the "interphase" stage. During the M phase, the cell undergoes nuclear division and cytokinesis. The interphase stage consists of the G1 phase in which the biosynthetic activities resume at a high rate after mitosis; the S phase in which DNA synthesis occurs and the G2 phase which continues until mitosis commences. While in the G1 phase, some cells appear to cease progressing through the division cycle; and are said to be in a "resting" or quiescent state (denoted as the "G0" state). Certain environmental factors (such as a lack of growth factors in serum or confluence of cell cultures) may cause cells to enter the quiescent state. Once the factor is restored, the cell should resume its normal progress through the cycle. However, cells grown in culture may be unable to enter the quiescent phase when the growth factor is removed, resulting in the death of these cells. This growth factor dependence is particularly relevant to cultured T cells. T lymphocytes that are exposed over a long term to high concentrations of IL-2 to promote cell growth often will die by a process called apoptosis if IL-2 is removed or if they are subsequently stimulated through the T cell receptor, i.e., if they encounter specific antigens. (see, e.g., Lenardo M. J., Nature, 353:858-861, 1991; Boehme S. A. and Lenardo M. J., Eur. J. Immunol., 23:1552-1560, 1992). Therefore, the culture methods used to propagate LAK cells or TIL-cells, and prior methods to culture T cells which predominantly rely on high long-term concentrations of IL-2 to promote expansion in vitro, may render many of the cells susceptible to apoptosis, thus limiting or eliminating their usefulness for cellular immunotherapy.
It may also be advantageous in cellular immunotherapy studies to use gene transfer methods to insert foreign DNA into the T cells to provide a genetic marker, to facilitate evaluation of in vivo migration and survival of transferred cells, or to confer functions that may improve the safety and efficacy of transferred T cells. An established method for stable gene transfer into mammalian cells is the use of amphotropic retroviral vectors (see, e.g., Miller A D, Current Topics in Microbiology and Immunology, 158:1-24, 1992). The stable integration of genes into the target cell using retrovirus vectors requires that the cell be actively cycling, specifically that these cells transit M phase of the cell cycle. Prior studies have introduced a marker gene into a small proportion of polyclonal T cells driven to proliferate with high doses of IL-2, and these cells were reinfused into humans as tumor therapy and provided a means of following the in vivo survival of transferred cells. (Rosenberg et al. New Engl. J. Med., 323:570-578, 1990). However, for human T cells (which cycle slowly when grown with standard techniques) the efficiency of stable gene transfer is very low, in the range of 0.1-1% of T cells. (Springett C M et al. J. Virology, 63:3865-69, 1989). Culture methods which more efficiently recruit the target T cells into the S and G2-M phases of the cell cycle may increase the efficiency of gene modification using retrovirus-mediated gene transfer (Roe T. et al., EMBO J, 2:2099-2108, 1993), thus improving the prospects for using genetically-modified T cells in cellular immunotherapy or using T cells to deliver defective genes in genetic deficiency diseases.
The rapid expansion method described by S. Riddell et al. (in PCT Publication WO 96/06929, published Mar. 7, 1996), hereinafter referred to as "high-PBMC REM" or "hp-REM" was developed to provide functional, antigen-specific T cell clones for use in clinical adoptive immunotherapy protocols. The hp-REM protocol was designed to provide maximal T cell expansion in a limited amount of time without loss of T cell function and specificity. Generally, the hp-REM protocol involves the steps of adding an initial T lymphocyte population to a culture medium in vitro; adding to the culture medium a disproportionately large number of non-dividing peripheral blood mononuclear cells ("PBMC") as feeder cells such that the resulting population of cells contains at least about 40 PBMC feeder cells (preferably at least about 200, more preferably at least about 400) for each T lymphocyte in the initial population to be expanded; and incubating the culture. In preferred embodiments of the hp-REM protocol, the T cells to be expanded are also exposed to a disproportionately large number of EBV-transformed lymphoblastoid cells ("LCL"), to an anti-CD3 monoclonal antibody (e.g., OKT3) (to activate the T cells via the T cell antigen receptor), and to the T cell growth factor interleukin-2 (IL-2).
In the hp-REM protocol, T cells are generally expanded using a vast excess of feeder cells consisting of peripheral blood mononuclear cells (PBMC) and possibly also EBV-transformed lymphoblastoid cells (EBV-LCL). T cells to be expanded typically represent less than about 0.2% of the cells in the hp-REM culture method. As described, the T cells can be activated through the T cell antigen receptor using an anti-CD3 monoclonal antibody (e.g. OKT3) and T cell proliferation can be induced using IL-2. Such hp-REM culture conditions were reported to result in a level of T cell expansion 100 to 200-fold greater than that reported by others.
However, for most uses, it would be preferable to avoid the use of large excesses of feeder cells (i.e. PBMC and EBV-LCL) in the preparation of T cells destined for clinical use. For example, PBMCs are derived from human blood and could represent a potential source of adventitious agents (e.g. human imunodeficiency virus, type 1 and 2; human T cell leukemia virus I, type 1 and 2; and hepatitis virus, such as hepatitis B, C and G), and EBV-LCL could represent a potential source of Epstein-Barr virus. In addition, the large-scale application of the hp-REM protocol would require a large supply of human peripheral blood to provide adequate numbers of feeder cells.
It would therefore be particularly advantageous to reduce the numbers of such feeder cells required or to replace them entirely. With these concerns in mind, the methods of the present invention (hereinafter referred to as "low-PBMC-REM" or "modified-REM") are designed to achieve rapid in vitro expansion of T cells without using the vast excess of PBMC and/or EBV-LCL feeder cells that are the key characteristic of the hp-REM protocol.